Harness the full power of Rawi by integrating it seamlessly into your terminal workflow with pipes, redirects, and automation.
Rawi is designed to be a true command-line citizen, integrating naturally with shell pipes, redirects, file operations, and automation scripts. This makes it incredibly powerful for developer workflows.
Send data directly to Rawi for analysis:
# Analyze code files
cat server.js | rawi ask "Review this Node.js code for security issues"
# Process command output
ls -la | rawi ask "Explain these file permissions"
# Analyze logs
tail -100 /var/log/app.log | rawi ask "Summarize these error patterns"
# Git workflow integration
git diff | rawi ask "Write a conventional commit message for these changes"# Analyze multiple files
cat src/*.js | rawi ask "Review this JavaScript codebase architecture"
# Process all Python files
find . -name "*.py" -exec cat {} \; | rawi ask "Suggest improvements for this Python project"
# Combine different file types
cat package.json README.md | rawi ask "Is the documentation consistent with the package configuration?"Save AI responses directly to files:
# Generate documentation
rawi ask "Create API documentation for a REST user service" > api-docs.md
# Create configuration files
rawi ask "Generate Dockerfile for Node.js app" > Dockerfile
# Generate code templates
rawi ask "Create React component template" > Component.tsx
# Write scripts
rawi ask "Create backup script for PostgreSQL" > backup.sh# Add to existing files
rawi ask "Add error handling section" >> documentation.md
# Build files incrementally
rawi ask "Add unit tests section" >> project-plan.md# Multi-stage data processing
curl -s api.example.com/data | \
jq '.' | \
rawi ask "Analyze this API response structure" | \
tee analysis.md
# Log analysis pipeline
grep ERROR /var/log/app.log | \
head -50 | \
rawi ask "Categorize these errors and suggest fixes" | \
mail -s "Error Analysis" admin@company.com# Comprehensive code review
git diff HEAD~1 | \
rawi ask --act code-reviewer "Review changes for quality, security, and performance" > \
review-$(date +%Y%m%d).md
# Security-focused review
find . -name "*.js" -newer yesterday.tmp -exec cat {} \; | \
rawi ask --act security-expert "Scan for security vulnerabilities"# Auto-generate documentation
for file in src/*.js; do
echo "## $(basename $file)" >> docs.md
cat "$file" | rawi ask "Document this JavaScript module" >> docs.md
echo "" >> docs.md
done# Add to your ~/.bashrc or ~/.zshrc
# Quick AI assistance
alias ai="rawi ask"
alias aic="rawi ask --act code-reviewer"
alias ais="rawi ask --act security-expert"
alias ail="rawi ask --profile local"
# File analysis
alias aifile="rawi ask 'Analyze this file:' <"
alias ailog="tail -100 | rawi ask 'Analyze these logs:'"
# Git integration
alias aicommit="git diff --cached | rawi ask 'Write a conventional commit message'"
alias aireview="git diff | rawi ask --act code-reviewer"# Smart AI helper function
ai_help() {
local context="$1"
shift
local question="$*"
case "$context" in
"code")
rawi ask --act software-engineer "$question"
;;
"debug")
rawi ask --act debugging-expert "$question"
;;
"security")
rawi ask --act security-expert "$question"
;;
"data")
rawi ask --act data-scientist "$question"
;;
*)
rawi ask "$context $question"
;;
esac
}
# Usage: ai_help code "How do I optimize this algorithm?"
# Usage: ai_help security "Is this authentication secure?"# Analyze any file with appropriate context
analyze_file() {
local file="$1"
local extension="${file##*.}"
case "$extension" in
"py")
cat "$file" | rawi ask --act python-expert "Review this Python code"
;;
"js"|"ts")
cat "$file" | rawi ask --act javascript-expert "Review this JavaScript/TypeScript code"
;;
"sql")
cat "$file" | rawi ask --act database-admin "Review this SQL code"
;;
"yml"|"yaml")
cat "$file" | rawi ask --act devops-engineer "Review this YAML configuration"
;;
*)
cat "$file" | rawi ask "Analyze this file"
;;
esac
}name: AI Code Review
on: [pull_request]
jobs:
ai-review:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v2
- name: Install Rawi
run: npm install -g rawi
- name: Configure Rawi
run: rawi configure --provider openai --api-key ${{ secrets.OPENAI_API_KEY }}
- name: AI Code Review
run: |
git diff origin/main...HEAD | \
rawi ask --act code-reviewer "Review this pull request" > \
ai-review.md
- name: Comment PR
uses: actions/github-script@v6
with:
script: |
const fs = require('fs');
const review = fs.readFileSync('ai-review.md', 'utf8');
github.rest.issues.createComment({
issue_number: context.issue.number,
owner: context.repo.owner,
repo: context.repo.repo,
body: `## 🤖 AI Code Review\n\n${review}`
});ai_review:
stage: review
script:
- npm install -g rawi
- rawi configure --provider anthropic --api-key $ANTHROPIC_API_KEY
- git diff $CI_MERGE_REQUEST_TARGET_BRANCH_NAME..HEAD | rawi ask --act code-reviewer "Review this merge request" > review.md
artifacts:
reports:
codequality: review.md
only:
- merge_requests#!/bin/bash
# daily-ai-helper.sh
echo "🤖 Daily AI Development Assistant"
echo "================================="
# Check git status
if git status --porcelain | grep -q .; then
echo "📝 Uncommitted changes detected:"
git status --porcelain | rawi ask "Suggest how to organize these changes"
echo ""
fi
# Review recent commits
echo "📊 Recent commit analysis:"
git log --oneline -5 | rawi ask "Analyze this recent development activity"
echo ""
# Check for TODOs
echo "📋 TODO analysis:"
grep -r "TODO\|FIXME\|HACK" src/ 2>/dev/null | \
head -10 | \
rawi ask "Prioritize these TODO items"#!/bin/bash
# generate-docs.sh
PROJECT_NAME=$(basename "$PWD")
OUTPUT_DIR="docs/generated"
mkdir -p "$OUTPUT_DIR"
echo "📚 Generating documentation for $PROJECT_NAME..."
# Project overview
echo "# $PROJECT_NAME Documentation" > "$OUTPUT_DIR/README.md"
echo "" >> "$OUTPUT_DIR/README.md"
# Architecture overview
find src -name "*.js" -o -name "*.ts" -o -name "*.py" | \
head -20 | \
xargs cat | \
rawi ask "Analyze this codebase and provide an architecture overview" >> \
"$OUTPUT_DIR/architecture.md"
# API documentation
if [[ -f "package.json" ]] || [[ -f "requirements.txt" ]]; then
find . -name "*route*" -o -name "*api*" -o -name "*endpoint*" | \
xargs cat 2>/dev/null | \
rawi ask "Generate API documentation from this code" >> \
"$OUTPUT_DIR/api.md"
fi
# Setup guide
cat package.json README.md 2>/dev/null | \
rawi ask "Create a setup and installation guide" >> \
"$OUTPUT_DIR/setup.md"
echo "✅ Documentation generated in $OUTPUT_DIR"#!/bin/bash
# smart-log-monitor.sh
LOG_FILE="/var/log/app.log"
LAST_CHECK_FILE="/tmp/last_log_check"
# Get timestamp of last check
if [[ -f "$LAST_CHECK_FILE" ]]; then
LAST_CHECK=$(cat "$LAST_CHECK_FILE")
else
LAST_CHECK=$(date -d "1 hour ago" +%s)
fi
# Get new log entries
CURRENT_TIME=$(date +%s)
NEW_ENTRIES=$(find "$LOG_FILE" -newermt "@$LAST_CHECK" -exec tail -100 {} \;)
if [[ -n "$NEW_ENTRIES" ]]; then
echo "🔍 Analyzing new log entries..."
echo "$NEW_ENTRIES" | \
rawi ask --act sysadmin "Analyze these log entries for issues and provide recommendations" | \
mail -s "Log Analysis $(date)" admin@company.com
fi
# Update last check time
echo "$CURRENT_TIME" > "$LAST_CHECK_FILE"# Add to ~/.bashrc
# AI-powered command suggestions
command_not_found_handle() {
echo "Command '$1' not found. Getting AI suggestion..."
rawi ask --profile local "I typed '$1' in bash but it wasn't found. What command did I probably mean?" | \
head -3
}
# AI-powered man page summaries
aman() {
man "$1" | rawi ask "Summarize this man page in simple terms with examples"
}
# Intelligent directory exploration
what_is_here() {
ls -la | rawi ask "Explain what's in this directory and suggest next steps"
}# Add to ~/.zshrc
# AI command explanation
explain_last_command() {
fc -ln -1 | rawi ask "Explain what this command does: $(fc -ln -1)"
}
# Bind to key
bindkey '^[e' explain_last_command
# Smart directory suggestions
suggest_commands() {
pwd | rawi ask "I'm in directory $(pwd). What useful commands could I run here?"
}# Add to ~/.config/fish/config.fish
# AI command suggestions
function ai_suggest
rawi ask --profile local "Suggest shell commands for: $argv"
end
# Explain command before running
function explain_and_run
echo "Command: $argv"
rawi ask --profile local "Explain what this command does: $argv"
read -p "Continue? [y/N] " confirm
if test "$confirm" = "y"
eval $argv
end
end# Analyze multiple configuration files
for config in *.conf *.ini *.yaml; do
[[ -f "$config" ]] || continue
echo "Analyzing $config..."
cat "$config" | rawi ask "Review this configuration file for best practices" > "analysis-$config.md"
done# Security scan across project
find . -name "*.js" -o -name "*.py" -o -name "*.php" | \
while read -r file; do
echo "Scanning $file..."
cat "$file" | rawi ask --act security-expert "Scan for security vulnerabilities" | \
grep -v "No issues found" >> security-report.md
done# Keep documentation in sync with code
git diff --name-only HEAD~1 | \
while read -r changed_file; do
if [[ "$changed_file" == *.js ]] || [[ "$changed_file" == *.py ]]; then
echo "Updating docs for $changed_file..."
cat "$changed_file" | \
rawi ask "Generate documentation updates for this changed code" >> \
"docs/updates-$(date +%Y%m%d).md"
fi
done# Good: Process data efficiently
curl -s api.example.com/data | \
jq '.results[]' | \
head -100 | \
rawi ask "Analyze this API data sample"
# Avoid: Processing too much data
# curl -s api.example.com/data | rawi ask "Analyze all this data"# Robust pipeline with error handling
analyze_logs() {
local log_file="$1"
if [[ ! -f "$log_file" ]]; then
echo "Error: Log file not found: $log_file" >&2
return 1
fi
tail -100 "$log_file" | \
rawi ask "Analyze these log entries" 2>/dev/null || \
echo "Error: Failed to analyze logs" >&2
}# Limit resource usage
large_file_analysis() {
local file="$1"
local lines="${2:-100}"
# Process only a sample for large files
head -n "$lines" "$file" | \
rawi ask "Analyze this file sample (first $lines lines)"
}Next Steps: